Rebuild failing due to wonky ipv6

AI-generated summary

The discussion revolves around a failing Discourse rebuild due to issues with IPv6. pfaffman initially encounters a problem while rebuilding with official plugins, resulting in a Gem::RemoteFetcher::UnknownHostError. Despite trying different versions, disabling plugins, and upgrading Docker, the issue persists. Stephen suggests troubleshooting by disabling official plugins and checking DNS resolution. pfaffman discovers that they can access the problematic URL within the existing container, but the rebuild still fails. Ultimately, pfaffman resolves the issue by disabling previously enabled IPv6 addresses, which were causing problems within Docker containers.

Trying to resolve my Patreon problem I’m doing a rebuild. I have only official plugins, and it’s failing with

I, [2019-05-19T22:35:13.979327 #15]  INFO -- : > cd /var/www/discourse && gem update bundler
ERROR:  While executing gem ... (Gem::RemoteFetcher::UnknownHostError)
    timed out (https://rubygems.org/specs.4.8.gz)
I, [2019-05-19T22:36:14.620134 #15]  INFO -- : 

I’m sorta stumped.

I also encountered the same problem.

if use verison: beta canot rebuild, only use tests-passed

Thanks, but I’m on tests-passed.

If you’ve upgraded to tests-passed you also can’t swap back to beta until beta catches up.

It can’t hurt to disable official plugins too, just to troubleshoot.

Good idea, but no joy. I removed all plugins, and it’s still failing with

I, [2019-05-19T23:53:44.891357 #14]  INFO -- : > cd /var/www/discourse && gem update bundler
ERROR:  While executing gem ... (Gem::RemoteFetcher::UnknownHostError)

I tried upgrading Docker and disabling IPv6. Still no joy.

Ok, and if you ping www.google.com does it resolve?

What result do you get if you nslookup ? Does the connection to DNS time out?

Well, in the existing container, I can curl https://rubygems.org/specs.4.8.gz without any problem.

I guess I’ll next see if I can crank up the latest discourse/base image and see what it’s doing.

Well, a while back I enabled ipv6, and something with that broke some stuff (but only some?) inside Docker containers. I turned off those IPv6 addresses and it seems that this and my other problem are resolved.